Feature sparse linalg solvers#2841

Open

abagusetty wants to merge 47 commits intoIntelPython:masterfrom

abagusetty:feature-sparse-linalg-solvers

Commits on Apr 3, 2026

Fix deprecated tol kwarg in SciPy host fallback
abagusetty
committed
sparse/linalg: complete LinearOperator algebra, CG/GMRES/MINRES with oneMKL hooks
abagusetty
committed

Commits on Apr 6, 2026

Add tests for scipy.sparse.linalg: LinearOperator, cg, gmres, minres
abagusetty
committed
Fix implicit numpy conversion; use .asnumpy() for dpnp arrays
abagusetty
committed
tests: add comprehensive sparse linalg tests for LinearOperator, cg, gmres, minres
abagusetty
committed
tests: rewrite sparse linalg tests to match dpnp test_linalg.py style
abagusetty
committed
Fix dtype inference: use int8 trial vector so matvec preserves operator dtype
abagusetty
committed
add onemkl sparse gemv pybind logic
abagusetty
committed
sparse: add pybind11 module, CMakeLists, and hook _sparse_impl into _iterative.py
abagusetty
committed
Remove redundant sparse_gemv.hpp passthrough header
abagusetty
committed
sparse: gemv.cpp includes gemv.hpp directly
abagusetty
committed
sparse: capture exec_q from CSR data at closure construction
abagusetty
committed
sparse/gemv: add missing headers, input validation, and MKL/SYCL exception handling
abagusetty
committed
sparse/gemv: replace explicit if/else type dispatch with 2-D dispatch table
abagusetty
committed

Commits on Apr 7, 2026

Commits on Apr 8, 2026

minor fixes
abagusetty
committed

Commits on Apr 13, 2026